Anton Helm
Verified Expert in Engineering
HPC Expert and Developer
Berlin, Germany
Toptal member since June 30, 2025
Anton is a high-performance computing (HPC) expert with hands-on experience optimizing applications on some of the world's largest supercomputers. With a robust DevOps background, he excels at managing scalable infrastructures. Anton's keen interest in AI technology fuels his expertise in high-performance computing solutions, system optimization, and contemporary deployment methodologies.
Portfolio
Experience
- High-performance Computing (HPC) - 15 years
- MPI - 15 years
- Python - 15 years
- Linux - 15 years
- Scientific Computing - 15 years
- C - 10 years
- Fortran - 10 years
- C++ - 3 years
Preferred Environment
Python, MacOS, C++, Fortran, C, Vim Text Editor, Linux, AWS Command Line Interface (CLI)
The most amazing...
...thing I've done is resolve a critical performance issue in an application while delivering a large data center, leading to millions of euros in cost savings.
Work Experience
HPC Application Analyst
ParTec
- Headed the technical development for NextSilicon, an innovative hardware platform designed for cutting-edge computing applications.
- Provided technical support for evaluating a new ARM CPU developed by SiPearl.
- Conducted the performance tuning and analysis of AI frameworks, including TensorFlow and PyTorch.
- Developed infrastructure for compiling and distributing a heterogeneous scientific software stack across cloud and on-premises hardware.
- Contributed to developing an HPC data center with cloud and on-premises components.
- Managed AWS infrastructure, including AWS IAM and AMI administration for the company and a key customer.
HPC Performance Engineer
Eviden
- Benchmarked and optimized applications for the latest-generation hardware to scale across 2,000+ nodes.
- Provided customer support by diagnosing significant performance degradation and resolving associated issues.
- Developed customized solutions for maintaining software stacks across multi-OS, multi-architecture, and multi-node HPC environments.
- Conducted benchmarking for procurement and acceptance testing of Europe's largest HPC data centers, including JUPITER Booster (JSC), Levante (DKRZ), Viper (MPCDF), Leonardo (CINECA), and MareNostrum 5 ACC (BSC).
Research Fellow
Instituto Superior Técnico
- Oversaw and directed extensive research endeavors aimed at developing advanced numerical models for precise simulations of large-scale physical problems.
- Established and maintained an HPC cluster on-premises with 600 nodes.
- Contributed to developing and implementing a virtual reality experience for scientific visualization, providing an immersive understanding of complex datasets.
Visiting Graduate Researcher
UCLA
- Developed a CI/CD system for an in-house particle-in-cell code using Buildbot, which is hosted on-premises.
- Supervised PhD students in integrating a novel parallelization scheme for load-balancing parallel particle-in-cell simulations.
- Benchmarked and optimized scientific codes to scale across hundreds of thousands of nodes on various HPC systems in the United States.
Research Fellow
Helmholtz-Zentrum Dresden-Rossendorf
- Developed a parallel post-processing pipeline in Python utilizing IPython Parallel, HDF5, and memory mapping (mmap) to process terabytes of data in seconds rather than hours.
- Implemented various laser pulse configurations into particle-in-cell simulation codes developed in Fortran, C++, and CUDA.
- Modeled particle acceleration processes in laser-plasma interactions using particle-in-cell simulations.
Experience
Tool for Building, Testing, and Hardening AWS AMIs
The Pytest framework ensures that one instance per AMI is created and subsequently terminated, guaranteeing that all AMI requirements are met while minimizing testing costs.
CLI to Manage Containerized Self-hosted GitHub Action Runner
https://github.com/ahelm/wusaThis tool was actively utilized to deploy runners across diverse hardware configurations and operating systems, enabling parallel execution of CI/CD pipelines for various in-house scientific applications.
Education
PhD in Physics
University of Lisbon - Lisbon, Portugal
Master's Degree in Physics
Dresden University of Technology - Dresden, Germany
Certifications
Deep Learning
DeepLearning.AI | via Coursera
Skills
Libraries/APIs
NumPy, MPI, OpenMP, Matplotlib, SciPy, HDF5, TensorFlow, PyTorch
Tools
Pytest, Vim Text Editor, Git, GitHub, AWS Command Line Interface (CLI), CMake, Ansible, Buildbot, OpenFOAM, Packer, AWS IAM
Languages
Python, Bash, Fortran, C, C++
Paradigms
High-performance Computing (HPC), Parallel Programming, Parallel Computing, DevOps
Platforms
Linux, MacOS, Jupyter Notebook, Docker, Amazon Web Services (AWS), NVIDIA CUDA
Other
Physics, Numerical Modeling, Plasma Physics, Scientific Computing, Performance Tuning, Spack, Profiling, Modeling, Applied Mathematics, Applied Physics, Cython, Slurm Workload Manager, Data Processing, CI/CD Pipelines, Analysis, GitHub Actions, Algorithms, Natural Language Processing (NLP), Supervised Learning, ROCm, JUBE, StarCCM+, Research, AWS Parallel Cluster, AWS AMI, InfiniBand, Cloud
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
Share your needs
Choose your talent
Start your risk-free talent trial
Top talent is in high demand.
Start hiring